ABSTRACT Glioblastoma cells trigger pharmacoresistant seizures that may promote tumor growth and diminish the quality of remaining life. To define the relationship between growth of glial tumors and their neuronal microenvironment, and to identify genomic biomarkers and mechanisms that may point to better prognosis and treatment of drug resistant epilepsy in brain cancer, we are analyzing a new generation of genetically defined CRISPR/in utero electroporation inborn glioblastoma (GBM) tumor models engineered in mice. The molecular pathophysiology of glioblastoma cells and surrounding neurons and untransformed astrocytes will be compared at serial stages of tumor development in three genetic mouse strains: wild type, seizure prone, and seizure resistant. Preliminary data reveal that epileptiform EEG spiking is a very early and reliable preclinical signature of GBM expansion preceding other neurological deficits in these mice, followed by rapidly progressive seizures and death within weeks. Transcriptomic analysis of cortical astrocytes reveals the expansion of a subgroup enriched in pro-synaptogenic genes that may drive hyperexcitability, a novel mechanism of epileptogenesis. In Specific Aim 1 we will systematically define the earliest appearance of cortical hyperexcitability in wild type mice with a prototypical GBM and correlate its progression with in vivo and neuropathological imaging of invasive tumor cell location, in vitro electrophysiology, and molecular markers of key epilepsy pathogenic cascades in peritumoral neurons, including impaired glutamate reuptake, altered GABA gated-chloride gradients, and synaptic densities. In Specific Aim 2 we will correlate these findings with detailed FACS-sorted transcriptomic profiles of both transformed and wild type astrocytes in the peritumoral region to test the novel hypothesis that peritumoral hyperexcitability is driven in part by astrocytic subtypes that disrupt synaptic E/I homeostasis. In Specific Aim 3, we will use this benchmark approach in WT brain to compare growth, electrophysiological and molecular pathological profiles of the same tumor generated in a hyperexcitable brain bearing a single gene deletion (Kcna1) that dramatically lowers the threshold for seizures and shortens lifespan, and in a monogenic deletion strain (MapT/tau) that raises cortical seizure threshold and prolongs life, in order to examine the contribution of host neuronal excitability to tumor expansion. Our approach sets the stage to broadly explore the developmental biology of personalized tumor/host interactions in mice engineered with novel human tumor mutations in specified glial cell lineages.